Pain is common in cancer patients. To ensure optimal pain management efficacy and effectiveness of new drugs and treatments have to be investigated in clinical trials. Efficacy trials such as randomised controlled trials (RCT) are experimental studies and estimate the maximum potential benefit to be derived from an intervention in ideal circumstances and under a controlled environment. RCTs are the only trial design to establish causal effects. A crossover study is a special type of RCT where patients serve as own controls. In efficacy studies the intervention and the control group should be as homogeneous as possible, confounding variables are controlled, bias is reduced, internal validity is high whereas external validity is low. Studies looking at effectiveness assess clinical practice and reflect real life circumstances. They rely high on external validity at the expense of careful controls, the study population is heterogeneous, confounding variables are examined. Cohort studies follow a group or groups of individuals with a common characteristic over a period of time to measure outcomes. Case-control studies start with the outcome and compare the characteristics of two groups of interest, those with the outcome and those without to identify factors which occur more or less often in the poor outcome group. Definition of outcome criteria is crucial both for efficacy and effectiveness studies and is often a primary problem. All clinical studies must use valid and reliable outcome measures.